DJCM: A Deep Joint Cascade Model for Singing Voice Separation and Vocal Pitch Estimation
The paper proposes a DeepJointCascade Model (DJCM) for singing voice separation and vocal pitch estimation tasks in music information retrieval. Traditional methods, classified into pipeline methods and naive joint learning…
Continue reading